Live Discussion: How the Other Half Lives—Or the What, How, and Where of the AβPP Intracellular Domain
Return to Discussion Text
Live discussion held 20 September 2002, featuring Frank LaFerla of the University of California, Irvine, and Paul Coleman (moderator) of the University of Rochester, NY.
Participants: Frank LaFerla, Paul Coleman (moderator), June Kinoshita, Jason Shepherd, Min Zhu, Stavros Therianos, Malcom Leissring, Keith Crutcher, Glenda Bishop, Andrea Wilson, Della. Stefan, Rong Wang.
Note: The transcript has been edited for clarity and accuracy.
Frank LaFerla: Hi, Paul.
Paul Coleman: Hi, Frank. I look forward to this.
Frank LaFerla: Me, too!
Paul Coleman: By the way-why are you represented by a fool's cap?
Frank LaFerla: I thought it was appropriate.
Paul Coleman: I disagree, but no matter. We are all in the same boat.
June Kinoshita has entered Alzheimer Research Forum
Jason Shepherd has entered Alzheimer Research Forum
June Kinoshita: Hi, Frank. Hi, Paul.
Paul Coleman: I am joined now by Min Zhu and Stavros Therianos. Stavros has a special interest in Notch, and Min has been working on nuclear localization of C-terminal APP.
Jason Shepherd Hi, Frank.
Frank LaFerla: Hi, Min, Stavros.
June Kinoshita: Frank, I invited Paul to moderate the discussion and direct the flow of questions. That will free me up to enforce some order and good manners!
Frank LaFerla: An excellent choice.
Andrea Wilson has entered Alzheimer Research Forum
June Kinoshita: Frank, I like your choice of screen icon.
Frank LaFerla: Thanks, I thought it was appropriate.
June Kinoshita: Are you the official jester in the kingdom of APP?
Frank LaFerla: Yes, the official fool.
Keith has entered Alzheimer Research Forum
June Kinoshita: I'd like to welcome everyone to today's chat. Could we begin with a round of quick self-introductions?
Glenda Bishop and Andrea Wilson: Hi, everyone. This is Glenda Bishop and a new lab colleague, Andrea Wilson.
Keith: Keith Crutcher here.
Malcolm Leissring has entered Alzheimer Research Forum
Paul Coleman: Frank, would you like to start with a brief overview of C-terminal in AD and its recent attention?
Frank LaFerla: Sure. I think the first description of this fragment was provided by Luciano D'Adamio's group several years ago. Unfortunately, their manuscript was published in the Journal of AD Research (sorry, no offense), so it was not widely seen by the field. (Editor's note: See our live discussion: "Generation of an apoptotic intracellular peptide by g-secretase cleavage of Alzheimer's b-amyloid precursor protein.")
Frank LaFerla: They also named the fragment AID for APP intracellular domain.
Paul Coleman: So, what gave it the current impetus?
Frank LaFerla: The field has now seemed to have adopted the name AICD because of similarities between it and Notch; the same fragment in Notch is referred to as NICD.
Frank LaFerla: Its impetus: Probably the most exciting aspect is that it is involved in nuclear signaling and, therefore, has the potential to amplify cellular responses.
Malcolm Leissring: Hasn't Rachael Neve been heralding the idea that the CTF of APP is the neurotoxic portion of APP?
Frank LaFerla: Yes, Rachael has been suggesting that it is neurotoxic, and several other groups as well, including Luciano D'Adamio's and Brad Hyman's.
Jason Shepherd: Perhaps it is neurotoxic through abnormal calcium signaling, as Frank LaFerla has shown?
Frank LaFerla: Good point, Jason.
Paul Coleman: It appears that Rachael is not here now, but it is correct that she championed the C100 molecule for some time.
Stefan has entered Alzheimer Research Forum
However, as Frank LaFerla points out in his background, the situation may be more complex than that. In fact, a recent Cell paper ( Baek et al.) suggests additional complexes.
Malcolm Leissring: So there is very little evidence for a ligand for APP at this moment-and this is not for lack of trying by many groups. Does it make sense, then, that the AICD should have some critical nuclear signaling role?
Frank LaFerla: Yes, it does.
Jason Shepherd: I would be surprised if it didn't, but the BACE1 knockouts seem to have no observable phenotype, although, of course, there may be compensation by other BACE-like enzymes.
Malcolm Leissring: What is the message that is being signaled?
Della: One paper showed Ab to bind to APP. Does this seem reasonable? (See Lorenzo, et al., Nature Neurosci Vol. 3 No. 5 May 2000, Amyloid b interacts with the amyloid precursor protein: a potential toxic mechanism in AD.)
Frank LaFerla: That remains to be worked out here.
Malcolm Leissring: I suppose it could be saying, "I am attached to the extracellular matrix" or some other, chronic signal.
Frank LaFerla: I suppose it could seem reasonable, but Ab is a notoriously sticky molecule, so the significance of its interaction would need to be worked out.
Malcolm Leissring: It would be really nifty if Ab itself had some signaling role.
Frank LaFerla: I agree, particularly if there was a disparity between Ab 40 and 42.
Malcolm Leissring: Or perhaps tragic for the therapeutic efforts aimed at removing Ab....
Jason Shepherd: If Ab does have a function, it doesn't seem to be essential for normal cognitive function...at least in mouse models.
Malcolm Leissring: But APP knockout mice have definite electrophysiological deficits.
Frank LaFerla: This is true, and BACE knockouts may also have similar deficits once they get measured.
Jason Shepherd: True, but APP itself has a host of other putative functions.
Glenda Bishop and Andrea Wilson: So far, the physiological function of Ab is completely unknown; no tests have really been performed to work that out.
Malcolm Leissring: And APPL knockout flies show interesting phenotypes that, interestingly enough, can be rescued by AICD. Or, I should say, overexpression in flies of constructs containing AICD produce interesting phenotypes.
Frank LaFerla: As do mice with APP, APLP1/2 knockouts.
June Kinoshita: Regarding the effects of APP knockout, I suppose one would have to work out the contributions of all of the different fragments to the electrophysiological deficit.
Glenda Bishop and Andrea Wilson: Does the AICD contain any part of the Ab sequence?
Frank LaFerla: Nope, it corresponds to the last 50 amino acids of APP.
Glenda Bishop and Andrea Wilson: Then, in some ways, the function of Ab is irrelevant when determining the effects of AICD.
Jason Shepherd: What do people think about Ab forming a calcium permeable pore?
June Kinoshita: Is that Lansbury's proposal?
Malcolm Leissring: That idea has been talked about and published on for years.
Frank LaFerla: Lansbury has been proposing this recently, but the original idea as far as I recall is from Arispe.
Frank LaFerla: I think there is more and more data supporting a role for a calcium or other ion-permeable pore.
Malcolm Leissring: There are lots of papers.
Della: If APP is involved in transport, and as Ab was proved to be present in the vesicles, could Ab be released as a signal to other neuRongs or to astrocytes?
Malcolm Leissring: That's an interesting idea, but the release would seem to occur at sites very distal from the nucleus, according to the evidence on the transport idea.
June Kinoshita: We're drifting a bit off-topic. Let's get back to AICD! (Ab gets plenty of attention already!)
Jason Shepherd: What are those results in flies, Malcolm?
Malcolm Leissring: Flies overexpressing constructs that contain an intact AICD show increased synaptic boutons at the neuromuscular junction. And flies overexpressing APP constructs in the wing show a "blistered wing" phenotype.
Frank LaFerla: AICD may be pleiotrophic.
Malcolm Leissring: These fly papers are very illuminating and I recommend them to everyone.
Jason Shepherd: References, Malcolm?
Malcolm Leissring: Kalpana White has done the most work in this area. Torroja et al., J Neurosci 19(18):7793 is one.
Della: Does anyone know what sort of DNA-binding site would be recognized by the C-terminus?
Frank LaFerla: Della: Some of this has already been worked out, but I suppose it would ultimately depend on what transcription factors it ultimately binds to and whether this is a cell-dependent effect.
Frank LaFerla: Yama Akbari in my lab will be giving a talk at SFN detailing at least one gene that is affected.
Jason Shepherd: I think recently it was found that AICD can bind to Src homology domains, which are very common motifs on other proteins.
Glenda Bishop and Andrea Wilson: Frank, what do you mean by AICD being pleiotrophic. What effects are you talking about?
Frank LaFerla: Glenda Bishop and Andrea Wilson: I mean that it may have multiple roles. Malcolm Leissring, when he was in my lab, showed that AICD could play a physiologic signaling role by modulating phosphoinositide calcium signaling. Other groups have shown that overexpression of AICD is neurotoxic and can induce apoptosis, although it is not yet clear if these are overexpression artifacts.
Glenda Bishop and Andrea Wilson:
Frank, sounds a lot like Ab. Some think it is good, some think it is bad....
Frank LaFerla: Glenda Bishop and Andrea Wilson: It could be that the signaling role, especially with regard to its function in modulating calcium signaling, underlies the effects on apoptosis.
Malcolm, can you go a bit more into why you think these fly papers are enlightening?
Well, the fly papers show the effects of overexpression of APP constructs in an intact, living organism. The take-home message seems to be that the AICD is a necessary domain for the effects. The effects seem to be of two types, depending on the locus of expression: (1) synaptic bouton differentiation and morphology, and (2) interaction with the extracellular matrix.
Glenda Bishop and Andrea Wilson: Has anyone tried overexpressing the AICD in mice yet?
Malcolm Leissring: I'm working on it. I have 12 founders!!
Frank LaFerla: You have 12 founders that overexpress AICD?
Malcolm Leissring: And Fe65, using a bicistronic construct.
Frank LaFerla: Cool-are you using an inducible approach?
Malcolm Leissring: In a way.... I'm using the CaMKII promoter, which does not come on until P14 or so.
Jason Shepherd: There seems to be evidence that AICD can be phosphorylated, which seems to regulate its binding to Fe65.... This would be interesting in terms of regulation.
Glenda Bishop and Andrea Wilson: Malcolm, do you have cell lines that overexpress AICD, too?
Jason Shepherd: Isn't AICD "naturally" overexpressed in cells with PS1 mutations?
Malcolm Leissring: Jason: NO, the ratio of Ab42 to total changes, not the overall amount of g-secretase. In fact, some data suggest that the amount of overall g-secretase activity actually goes down for PS mutants, at least for Notch proteolysis.
Della: I would like to know, as Mark Bothwell seems to think, if the C-term really has a PEST sequence? As you could mutate this to stabilize the C-term in the transgenic mice or cells.
Rong has entered Alzheimer Research Forum
Jason Shepherd: I can dig out the reference for you, Della.
Malcolm Leissring: Della, what is a PEST sequence?
Della: Malcolm Leissring, this is a signal to the proteasome system for degradation.
Frank LaFerla: It is a degradation sequence.
Rong: Hi, June. How are you? I am Rongg Wang (Mt. Sinai School of Medicine).
Malcolm Leissring: AICD gets degraded by insulin-degrading enzyme-NOT by the proteosome.
Frank LaFerla: Christian Haas's group recently showed that AICD is degraded by IDE.
Della: AICD has been shown last year to be degraded in vitro by the proteasome: Eur. J Biochem 268:5329-5336, The C-term fragment of the AD amyloid protein precursor is degraded by a proteasome-dependent mechanism distinct from g-secretase.
Frank LaFerla: By the way, getting back to the AICD story, Luciano's group has a paper that is out in JBC showing that APLP1 and 2 also release an AICD-like domain. (See Scheinfeld, et al.)
Malcolm Leissring: No big surprise.
Glenda Bishop and Andrea Wilson: Is the AICD-like domain from APLP molecules very similar to AICD from APP?
Malcolm Leissring: Glenda and Andrea-yes. Almost totally identical.
Frank LaFerla: APLPs lack the Ab sequence, but the AICD are very similar.
Malcolm Leissring: All contain a GYENPTY sequence.
Glenda Bishop and Andrea Wilson: Malcolm, is that sequence the "active" part?
Malcolm Leissring: Glenda and Andrea, GYENTPY is a domain that is involved in internalization and interaction with other proteins like Fe65. Whether these two functions are related is not known.
Jason Shepherd: Does overexpression of APP itself lead to increased AICD, though?
Malcolm Leissring: Jason-it should, which makes the reports that AICD is toxic quite curious.
Frank LaFerla: Jason: It may be that AICD levels are very tightly regulated and that overexpressing APP will not lead to increased AICD. Christian Haas's/Steiner's group has shown that presenilin mutations which increase Ab42 levels do not necessarily lead to a concomitant increase in AICD.
Malcolm Leissring: Frank, but if Ab goes up (as it does), then by definition AICD also goes up. I suppose that the amount of "signaling-competent" AICD might be regulated, though.
Frank LaFerla: Malcolm, the processing may go up, but the steady levels may not change.
Jason Shepherd: Perhaps the AICD-like fragments from APLPs compensate for the function of AICD in the BACE1 KO and other models where AICD production is inhibited...thus, the actual function of AICD may be masked.
Rong: Frank, I got here late. Can you update for me on the relationship between the generation of Ab and AICD? Are they the same processing event or independent from each other? Thanks.
Frank LaFerla: Rong: It appears as though AICD is liberated by a g-secretase type activity that occurs at a downstream site that is referred to as the epsilon cleavage site.
Frank LaFerla: Cleavage at this epsilon site is presenilin-dependent.
Rong: Frank, so the generation of Ab and generation of AICD are independent proteolytic events, right?
Frank LaFerla: Rong, there are three papers (Sisodia's, Beyreuther, and Haas's groups) showing that AICD is cleaved downstream of the Ab sequence. I can e-mail you the references if you want them.
Malcolm Leissring: I've got to run, everyone. Thanks for chatting on this exciting topic!!
June Kinoshita: Frank, would the AICD be different depending on whether you have Ab40 or 42?
Frank LaFerla: June: Good question. Since cleavage occurs downstream at position 49 or 50 (relative to Ab), it would be expected that AICD is not that different. Otherwise we would have predicted that an AICD 57 or 59 would have emerged depending on whether Ab42 or 40 was generated.
Frank LaFerla: June, that doesn't mean that AICD 57 or 59 doesn't exist. Luciano's group did some mass spect analysis and showed that a small amount can be found in vivo.
June Kinoshita: Interesting, Frank. And there have been reports of elevated C99 in AD brain. So do those brains have more or less AICD? Also, is it possible that there might be functional differences between the AICD 57 and 59 fragments?
Frank LaFerla: June, as far as I know, no one has quantified AICD 57 or 59 levels in different stages of AD; the closest report was by Luciano -- his group in their original paper published in J. AD Research.
June Kinoshita: Rong, is that something you're looking at (quantifying AICD 57 and 59) in AD brain?
Rong: June, not yet.
Glenda Bishop and Andrea Wilson: Can the AICD fragment be cleaved from APP while Ab is still attached to the N-terminal region, or can it only be cleaved after Ab is removed?
Frank LaFerla: Andrea, according to those three reports, it looks like AICD is liberated first, followed by cleavage at the g-secretase site.
Glenda Bishop and Andrea Wilson: Will AICD cleavage always then lead to g-cleavage?
Frank LaFerla: Curiously, it seems like the b-CTF (C99) is much more a substrate for generating AICD than the a-CTF (C83).
Glenda Bishop and Andrea Wilson: So can AICD be cleaved from the full APP as well as from CTF?
Frank LaFerla: I think it needs to be cleaved from C99. Not sure if it can be generated from the holoprotein. It might not be a good substrate.
Jason Shepherd: It was in a recent JBC paper...can't remember the authors offhand, though. I'll see if I can dig it out.
Paul Coleman: Hi! I am Stavros, working in Paul's lab. What is known about AICD transport into the nucleus? Could this be a way of regulating the DNA-binding abilities of APP-C?
Paul Coleman: Stavros writes: Would different levels of AICD 57 and 59 have a differential effect on Fe65 nuclear localization?
Frank LaFerla: Stavros: I don't think that has been addressed yet. They are differentially unstable, at least in cell culture.
June Kinoshita: We have five minutes to wrap up. Frank LaFerla, what are the most pressing open questions with respect to AICD that you would like to see answered?
Frank LaFerla: 1. What genes are specifically upregulated by AICD?
2. Do the APLP1 and 2 AICD-like domains compensate for APP AICD?
3. I would like to see AICD be overexpressed in a mouse model; I think it would have to be done in an inducible fashion.
4. What effect does AICD have on the pathogenesis of AD? Is it a normal physiologic function, or does it truly lead to a pathological consequence, as well (perhaps apoptosis as some groups have suggested)?
I think those are some of the important questions that need to be addressed.
Jason Shepherd: Microarray would be interesting to determine what genes are turned on/off.
Frank LaFerla: Bruce Yankner reported at the International Conference that he gene-arrayed following AICD expression. (Editor's note: See our related news story .)
Frank, I suppose you'll be calling Malcolm about his mice!
Frank LaFerla: June, the last thing I need is more mice. We have about 5,000 here in my colony!
June Kinoshita: If you have mice to give away, you can post on the website!
Della: Jason, you said that you had a reference concerning the PEST sequence in AICD. That would be great. My e-mail is email@example.com Thanks!
June Kinoshita: Frank, can you give us a preview of what you will be presenting at Neuroscience?
Frank LaFerla: At SFN: Yama Akbari will be talking about a gene candidate that is selectively upregulated following AICD treatment, and Salvo Oddo will present a poster on a new triple-transgenic mouse model that develops Ab and tau pathology as well as deficits in synaptic plasticity.
Frank LaFerla: As for the AICD story, an article that I wrote for Nature Reviews Neuroscience on this topic will appear in their November issue and also be distributed at the conference.
June Kinoshita: Great, I'll keep an eye out for that issue.
Frank LaFerla: Okay, great.
June Kinoshita: I've thought about starting an APP C-terminus club for people with an interest in the various C-terminus fragments. You could hold regular chats or threaded discussions. Would that be interesting to all of you?
Frank LaFerla: Might be interesting.
Della: Could you also do that for other topics like tau?
Frank LaFerla: Thanks for the chat. It was fun.
June Kinoshita: I'd also be interested in hearing from folks about the availability of good reagents to study non-Ab products of APP processing.
Stefan: As long as it doesn't conflict with my APP N-terminus club meetings....
June Kinoshita: Glenda and Andrea, sure, we can start other clubs.
Paul Coleman: June, your proposed club would be of interest to us.
Frank LaFerla: Look forward to seeing you all at Neuroscience in Orlando.
June Kinoshita: Frank, thanks so much for leading today's discussion. See you in Orlando!